ARTICLE IN PRESS

Energy Policy 32 (2004) 20672076

Electricity theft: a comparative analysis

Thomas B. Smith*
Department of Social and Behavioral Sciences, Zayed University, P.O. Box 19282, Dubai, United Arab Emirates

Abstract
Electricity theft can be in the form of fraud (meter tampering), stealing (illegal connections), billing irregularities, and unpaid bills.
Estimates of the extent of electricity theft in a sample of 102 countries for 1980 and 2000 are undertaken. The evidence shows that
theft is increasing in most regions of the world. The nancial impacts of theft are reduced income from the sale of electricity and the
necessity to charge more to consumers. Electricity theft is closely related to governance indicators, with higher levels of theft in
countries without effective accountability, political instability, low government effectiveness and high levels of corruption.
Electricity theft can be reduced by applying technical solutions such as tamper-proof meters, managerial methods such as inspection
and monitoring, and in some cases restructuring power systems ownership and regulation.
r 2003 Elsevier Ltd. All rights reserved.
Keywords: Electricity theft; Governance; Corruption

1. The Problem
An electric power system can never be 100% secure
from theft. In many systems the amount of theft is small
(12%) in terms of the electricity generated. But, the
nancial loss is high due to the large amount of
electricity distributed. Nesbit (2000) noted that, In
the US, the consensus seems to be that theft costs
between 0.5% and 3.5% of annual gross revenues in the
US. That seems like a small amountuntil you consider
that US electricity revenues were in the $280 billion
range in 1998. Therefore, between $1 and $10 billion
worth of electricity was stolen.
Some power systems may forfeit more than 15% of
power generated to various types of theft. A Transparency International (1999) report explains the situation in
Bangladesh.
In scal 199899 Bangladesh Power Development
Board (BPDB) generated 14,150 MkWh of electricity,
purchased another 450 MkWh from private sources,
but billed for only 11,462 MkWh, giving a system loss
of 22%. This was better than Dacca Electric Supply
Authority (DESA) 40% but poorer than Rural
Electrication Board (REB) 17%. The weighted
average system loss in the power sector as a whole
*Tel.:+971-4-2082-540; fax: +971-4-3522-534.
E-mail address: tbsmith9@yahoo.com (T.B. Smith).
0301-4215/03/$ - see front matter r 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0301-4215(03)00182-4

is estimated at 35%, which includes 21% technical

loss. The balance 14% y was due to pilferage, theft
and unauthorized use.
The nancial losses are critical to many electric power
organizations. Lost earnings can result in lack of prots,
shortage of funds for investment in power system
capacity and improvement, and a necessity to expand
generating capacity to cope with the power losses. Some
power systems in worst affected countries are near
bankrupt. Corruption increases and becomes entrenched as favors can be bought from power sector
employees in the form of inaccurate billing and allowing
illegal connections. Political leaders intervene to ensure
that cronies and supporters are not prosecuted. In 1998,
the situation deteriorated in Pakistan to the extent that,
The government took action and employed 35,000
army men to recover Water and Power Development
Authority (WAPDA) dues and curb the theft. They
have been conducting house-to-house raids with the
staff of WAPDA, checking for any tampering of
power meters. In the last year the army has found
100,993 instances of power theft, recovered Rs. 2.4
billion in nes and penalties and arrested 1188
people. Embarrassingly, many of the thefts were
discovered in the houses, farms and mills of the ruling
party legislators, 13 of whom were WAPDA ofcials.

ARTICLE IN PRESS
2068

T.B. Smith / Energy Policy 32 (2004) 20672076

Even the Minister for Population y resigned from

her cabinet post on power theft charges (Rizvi, 2000).
Electricity theft is a complex phenomenon with many
facets. In this article, electricity theft is dened and
various types of theft are described. The international
scope and trends of theft will be examined. How theft
can become institutionalized as part of the political,
economic and managerial culture of governance will be
noted. Lastly, some methods of dealing with the
problem of electricity theft are examined.

2. An increasing awareness
The emergence of power theft as a serious problem
has evolved due to several recent trends. Most countries
developed electric power systems that were highly
centralized state owned monopolies, where efciency
and prots were not high priority. The privatization of
the infrastructure and new modes of power policy
requires the new business-like enterprises to operate
efciently and try to optimize prots in an environment
of rapid change (Flavin and Lenssen, 1994; Patterson,
2001).
In many countries power theft is an issue of open
discussioneven in the most efcient (such as in the
USA) and moderately efcient (Malaysia) systems. In
South Asian countries, electric power is rarely discussed
without reference to power theft, since it is such a
prevalent practice. However, in some countries (Thailand, China) the topic is rarely part the analysis of
power systems.
The World Bank and Transparency International
have been major forces in promoting efcient, noncorrupt governance of public and private sector
organizations that provide essential services to the
public. The development of a civil society in many
countries has lead to a questioning of the effectiveness
and efciency of infrastructure and service delivery and
promotes transparency in the operations of secretive
bureaucratic organizations.

3. Inefciency in electric power systems

In power systems, not all electricity generated reaches
paying consumers because of inefciencies and managerial practices.
3.1. System use
The electric power system will consume some of the
electricity generated in order to produce and distribute
the electricity. Lighting, maintenance, etc. have to be
provided in power stations and facilities and this power

is not subject to metering or sold in the commercial

sense. System use can vary from 2% to 6% of
generation.
3.2. Technical transmission and distribution (T&D)
losses
Electric power is lost while being transmitted and
distributed when it passes through transformers and is
distributed in small capacity lines. Systems with long
transmission lines risk a higher amount of T&D loss
than shorter line systems. Also, the quality of the lines
and transformers can affect efciency of transmission
and distribution. Power systems that invest signicantly
in maintenance and in the advanced technology of
transmission and distribution can reduce technical T&D
losses.
3.3. Gratis
A third type of loss is electricity allocated gratis. Some
power systems provide free electricity without charge to
certain people and organizations. The presidential or
prime-ministers residence, members of parliament,
or the royal palaces may not be charged for electricity
or telephone. Free electricity may be provided to
employees of the power system as in Thailand where
the 32,000 employees of the Electricity Generating
Authority of Thailand (EGAT) receive free electricity
worth 1.5 billion baht.

4. Dening electricity theft

Four kinds of theft are prevalent in all power
systems. The extent of the theft will depend upon a
variety of factorsfrom cultural to how the power
utility is managed.
4.1. Fraud
Fraud is when the consumer deliberately tries to
deceive the utility. A common practice is to tamper with
the meter so that a lower reading of power use is shown
than is the case. This can be a risky procedure for
an amateur, and many cases of electrocution have
been reported. In Malaysia, professionals have
approached residents and managers of businesses
offering to x the meter for a moderate fee (New
Straits Times, 1999). During 2 months of raids in
Malaysia on suspected areas 587 (86%) out of 684
inspected were conrmed to have tampered with their
meters or stolen electricity (The Star, 1998). The losses
can be substantial when fraud is by large organizations.
In Aurangabad, India,

ARTICLE IN PRESS
T.B. Smith / Energy Policy 32 (2004) 20672076

The 22 proprietors of Jalnas seven mini-steel plants

accused of massive power theft detected in Mondays
raids by the Maharashtra State Electricity Board
(MSEB) are absconding following the rejection of
their plea for anticipatory bail by the Sessions Court
y The MSEB has conclusive proof of the Rs 20crore (Rs 200 million) power theft y (including)
extremely sophisticated equipment the steel plants
used to doctor their electricity metersy(Indian
Express, 1998).
4.2. Stealing electricity
Electricity theft can be arranged by rigging a line from
the power source to where it is needed bypassing a
meter. In South Asian countries this practice is quite
common in poor residential areas where those wanting
electricity may not have lines allocated and may not be
able to pay if they were connected. Called the kunda
system in Pakistan, this practice is often accepted by
power managers as a fact of life in poor communities. In
Soweto, South Africa 6 tons of spiderweb cable used
for such connections was recovered in 6 months by the
electrical authority in raids (Campbell, 1999).
In Mexico,
The millions of illegal customers, who steal electricity
with wires known as diablitos, or little devils, have
pushed an overburdened electrical grid over the edge.
ythousands of homes and businesses have been hit
with power outages that electric company ofcials
blame largely on pirates. Published reports say the
thefts result in the loss of as much as $475 million
revenues annually (Sullivan, 2002).
The illegal lines are easy to detect as they are often
above ground and highly visible. However, one nds
reports of staff being assaulted and needing police
security to carry out the removal of the lines. Corrupt
staff from the electricity organization may take bribes to
allow the practice to continue. On a larger scale,
businesses may bribe power organization staff to rig
direct lines to their buildings or ofces and the power
does not go through a meter. The bribes can be much
less than the cost of the power. Money also can be given
to inspectors to keep them from nding and/or
reporting the theft.
4.3. Billing irregularities
Billing irregularities can occur from several sources.
Some power organizations may not be very effective in
measuring the amount of electricity used and unintentionally can give a higher or lower gure than the
accurate one. The unintentional irregularities may even
out over time. However, it is also very easy in some
systems to arrange for much lower bills to be given than

2069

for the power actually used. Employees may be bribed

to record the meter at a lower number than is shown.
The consumer pays the lower bill and the meter-reader
earns unofcial salary. In another type of billing
irregularity, ofce staff can move the decimal point to
the left on the bill so that a person or company pays
$47.48 instead of $474.80. Consumers may know that
some power organization staff are on-the-take for
providing these services. Employees may keep payments. A scheme in operation in Malaysia in the late
1990s diverted $1.59 million to private accounts before
detection (BRDC, 2000). The staff can easily earn from
this type of corruption, as it is not easy to detect.
Corrupt practices may become institutionalized to the
extent that employees regard the illicit payments as part
of the job.
4.4. Unpaid bills
Some persons and organizations do not pay what they
owe for electricity. Residential or business consumers
may have left the city or an enterprise has gone
bankrupt. In South Africa, a culture of non-payment
is evident (Mkhwanazi, 1999). In Armenia, Nonpayment levels of 8090% are typical in the residential
sector. T&D losses are over 40% (Tacis, 1998). The
practice is widespread,
Some systems have chronic non-payersthe very rich
and politically powerful who know that their
electricity will not be cut regardless of whether they
pay or not. In India, farmers in some states regard
electricity as a free service from government, and
some political leaders and parties curry favor by
promoting this practice and prevent the State
Electricity Boards from collecting. Another chronic
non-payment group can be government departments
and agencies. The Pakistan Army discovered that
some of the largest amounts owed to WAPDA were
from government agenciesincluding the Army
itself. The Karachi Electric Supply Corporation
Director reported in 2000 that only 52 percent of
the 1.67 million customers were paying their bills
(News International, 2000).
In Indonesia in 2000, the military owed Rp. 23 billion
(US$3.1 million) to Perusahaan Listrik Negara (PLN).
This was a large part of the companys total unpaid
claims of about Rp. 157 billion (Jakarta Post, 21 March
2000). Some analysts may not regard non-payment by as
theft. However, when it becomes institutionalized and
people and organizations expect that they can get away
with it, unpaid bills should fall into the theft category.
Non-payment is a problem not conned to poor
countries. Lundin (2001) has explained the growing
problem in the USA. In all countries, as electricity
increases in price, some people have trouble paying their

ARTICLE IN PRESS
2070

T.B. Smith / Energy Policy 32 (2004) 20672076

bills regularly. This may encourage them to nd ways of

reducing their bills, such as tampering with the meter.
In a more conventional denition of electricity theft
the category of Unpaid Bills may not appear. However,
in some power systems the extent of the problem and its
impact has serious consequences. Data on non-payment
is not available easily that can be used for a comparative
analysis for the purposes of this paper. The analysis in
this paper deals primarily with theft in terms of billing
irregularities, fraud and stolen electricity.

5. Measuring electricity theft

Electricity theft can be estimated, but not measured
exactly. The most accurate estimate of theft is by
conducting a thorough analysis of the power system.
The revenue protection section of the Arizona Public
Service Company (APS) carried out a recent study that
is unique (Culwell, 2001). The APS provides electric
power to the Phoenix metropolitan region and 11 of
Arizonas countiescovering 40,000 miles2 with 868,000
customers. The APS wanted a research project that
would go beyond the usual studies that target meter
tampering. They wanted to know the extent of meter
tampering and the nancial loss in such a way as to be
able to extend the research to the whole of the APS
system with a 95% condence. The study involved
selecting randomly 550 meters out of the 868,000,
ensuring that they were spread among the urban and
rural users (35% rural) and residential and industrial
(12%) users. Each meter was thoroughly inspected
disconnected, opened, tested, and 52 items of information recorded about the meter. For determining theft the
beyond a reasonable doubt criterion was used.
Suspected theft required evidence that was clear and
convincing. The research study was implemented
beginning on 3 April 2000 and was completed by 30
June the same year. The ndings include:
*
*
*

Denite meter tampering rate0.72%.

Probable meter tampering rate1.00%.
Actual loss in dollars$330,148.

The data was extrapolated to the APS system to

estimate that nearly 15,000 meters had been tampered
with and show that the tampering losses per year were
estimated to be $7,967,279 that was 0.518% of revenue
loss for the APS. The APS study noted that the
estimated loss ($5.1 million) was much higher among
commercial accounts than the residential consumers.
The standard method of measuring power theft is by
analysis of transmission and distribution losses (T&D
losses). The method takes the difference between the
amount of electricity generated (minus system use and
gratis) in relationship to the amount metered and sold.
If an accurate calculation is made of technical line

losses, theft may compose a large part of the unaccounted amountthe non-technical line losses in the
distribution network.
Very efcient power systems have less than 6% T&D
lossestheft may be 12%. Less efcient systems may
have 912% T&D loss and inefcient systems have line
losses of over 15%. The Malaysian Tenaga system has
T&D losses of 11% that includes theft losses estimated
at 4%. Bangladesh estimates are T&D losses of 35%
with 14% as theft. In Budapest, Elmu estimates that half
of its 13% losses are due to theft (East European Energy
Report, 1999). Indonesias PLN estimated theft in
power distribution in Jakarta at 7% in 1994 and
3.77% in 1996 (Priatna, 1999). Thus, a system operating
with 22% T&D losses could lead analysts to estimate
that around 1015% are due technical T&D losses. The
remaining 712% of the electricity disappeared, probably due to theft of various types. This is a blunt method
for estimating theft and does not include non-payment.

6. Power theft: a comparative and historical perspective

Information is available on T&D losses for many
countries from the World Bank. However, World Bank
data on T&D losses for some countries is inaccurate and
misleading as 0 T&D losses are recorded, or the gure
given is less than 1%. This is impossible because some
electricity always is lost during transmission and
distribution.
It is neither realistic nor feasible to assess T&D losses
in all countries given the limitations in the data. For this
study, a sample of 102 countries was chosen. The basic
data for the countries is from the World Banks
Development Indicators (2003). The main criteria for
selection are:
*

Available data on T&D losses for 1980 and 2000 to

enable an historical perspective.
Reasonable condence in the accuracy of the data
and that system use was not included.
Countries selected have a good record in the
collection of data in other social, economic and
power sector variables.
The conrmation of the country data by a second
source such as the US EIA, reports on energy
development, and statistics bureaus and electricity
organizations in the selected countries.

Table 1 shows the distribution of T&D losses within

ranges. If 16% or more T&D losses are viewed as
systems with extensive theft, then over 40% of the
sample countries have a serious problem.
Table 2 provides an overview of T&D losses by region
for 1980 and 2000. The bottom row of the table shows
that between 1980 and 2002 the losses increased from
11.69% to 16.23%. Comparing regional losses reveals

ARTICLE IN PRESS
T.B. Smith / Energy Policy 32 (2004) 20672076
Table 1
T&D losses in 2000
T&D losses

Number of countries

Percent

410%
1115%
1620%
2153%
Total

34
25
15
28
102

33.3
24.5
14.7
27.5

Table 2
T&D Losses by region, 1980 and 2000
Region

Countries 1980
T&D
loss
percent

Western Europe
17
Eastern Europe
24
Middle-east, North Africa
11
Africa
11
North America
3
South America
9
Central America, Caribbean
9
South Asia
5
Southeast Asia
7
East Asia, Australasia
6
Total
102

7.71
9.68
11.18
14.6
9.67
13.00
15.50
25.20
12.14
8.67
11.69%

2000
Change
T&D
percent
loss
percent
7.56
18.18
19.63
19.95
9.38
17.23
21.68
27.55
13.32
7.65
16.22%

0.15
+8.50
+8.45
+5.35
0.29
+4.23
+6.18
+2.35
+1.18
1.02
+4.54%

that in only two regions did T&D losses decline in the

20-year period under reviewWestern Europe and
North America. South Asia in 1980 had the highest
T&D losses among the regions and further declined in
the 20-year period. The highest increase appears in the
Middle East/North Africa Region and in Eastern
Europe.
The disturbing evidence is that losses (and theft)
appear to be increasing in an era of readily available
technological means (metering, for instance) to lower
non-technical losses. Also, many of the power systems
were privatized between 1980 and 2000, yet the data
shows that overall, this has not impacted positively on
power sector efciency.
The lowest T&D losses (less than 6%) are in countries
known for efciency in management such as Finland,
Germany, Japan, Republic of Korea, Netherlands,
Singapore, Belgium, Austria, France and Switzerland.
The power organizations are managed to ensure the
deterrence, detection and prosecution of people and
organizations engaged in electricity theft. While there is
a low percentage of theft, the economic losses can be
high due to the large amount of electricity generated.
High losses (over 30%) are in countries such as
Albania, Haiti, Myanmar, Kyrgyz Republic, Nigeria,
and Bangladesh. Common features are poverty and that
each country has experienced political, economic and

2071

social turmoil. In tumultuous times government organizations cease to function efciently, become prone to
corrupt practices, investment is not made in system
management, and the consumers take advantage of the
system
Variations in T&D losses within each country may be
large. In the Philippines the T&D losses were estimated
to be 17% in 1997. However, assessment of regional
variations shows that six of 15 regions had losses below
17%. One region has over 27% loss and ve were
between 20% and 27%. The Meralco region (Manila)
reported losses of 12.4%, well below the rural areas
(National Economic Development Authority, 1998,
Table 5.4).
India has overall T&D losses of over 26%, but the
losses vary in the 22 states. Losses of nearly 50% are
experienced in Delhi, Jammu and Kashmir, and Orissa.
Even Maharashtra, with the best record, has nearly 15%
losses.

7. Governance and electricity theft

Understanding governance has emerged as an important element in explaining patterns of social,
economic and political development Kaufmann et al.,
1999). Electricity theft is related to a broader culture of
governance or mal-governance. The World Bank
Institutes Governance, Regulation and Finance Unit
have compiled useful data. Attempting to measure
governance, Kaufmann and associates developed six
measures to assess the various dimensions of governance. Multiple indicators were used to measure each
dimension for 175 countries. The dimensions are:
Voice and accountability: Aspects of the political
process, civil liberties and political rights.
Political instability and violence: The likelihood that
the government may be overthrown by violent means.
Government effectiveness: The quality of public service
provision and the bureaucracy, competence of civil
servants and the independence of the civil service from
political pressure.
Regulatory burden: Incidence of market un-friendly
policies such as price controls, and perceptions of
burdens imposed by excessive regulation.
Rule of law: Abiding by the rules of society,
effectiveness of the judiciary, and enforceability of
contacts.
Graft and corruption: The exercise of public power for
personal gain, bribery, impact of corruption on business.
Table 3 provides the correlation analysis of T&D
losses (2000) for the 102 countries of the sample of this
paper with the data from the Governance project
created by Kaufmann et al. (2003). A negative correlation indicates that as the country scores on the

ARTICLE IN PRESS
T.B. Smith / Energy Policy 32 (2004) 20672076

2072
Table 3
Governance indicators and T&D losses
Governance
dimensiona
Voice and accountability
Political instability and violence
Government effectiveness
Regulatory burden
Rule of law
Graft/corruption

Correlation
T&D losses
0.496
0.555
0.682
0.503
0.644
0.636

Level of
signicance
>0.000
>0.000
>0.000
>0.000
>0.000
>0.000

In the Indian, Pakistan and Bangladesh cases, the

overwhelming evidence is that corrupt practices are
widespread in the electricity sector. The Lucknow
Electrical Services Authority General Manager conceded that, Out of 110 millions unit of electricity
supplied to the residents of Lucknow, at least 33% are
pilfered and resulting in losses worth Rs 100 crores (Rs 1
billion) every year. He also admitted that most of the
pilferage took place in connivance with power employees (Tripathi, 2000).

N=102. Pearsons r correlation coefcient.

a
Data from Kaufmann et al. (2003).

8. The consequences of electricity theft

Table 4
Total governance scores and T&D Losses
Total governance
quartiles

Mean T&D
loss percentage

Number of
countries

Highest governance quartile

Second highest governance quartile
Second lowest governance quartile
Lowest governance quartile
Total

7.57
13.33
22.00
22.51
16.22

24
26
26
26
102

governance indicator go lowerindicating poor governance, one will nd higher T&D losses (and theft).
The T&D losses are highly correlated with each of the
governance dimension and all correlation coefcients
were signicant at less then the 0.001 level. Countries
with higher T&D losses lack of civil rights, democratic
institutions and accountability and endure political
instability and possible violence. Government effectiveness is weak and there is a high regulatory burden. Rule
of law is weak. The relationship of T&D losses with
graft and corruption indicates that the two are closely
intertwined. Power theft thrives (and is a part of)
systems of mal-governance. The reverse can be said also
to be true, that in well-governed systems, one nds much
less power theft. The relationship of power theft to graft
and corruption is not unexpected as examples of high
power theft systems show evidence of corrupt practices
within power sector organizations.
In order to illustrate the relationship between
governance and T&D losses, the six governance
indicators were combined into a total governance score
for each country. The countries were then divided into
quartiles on the basis of the total governance score.
Table 4 provides the mean T&D losses for the countries
in each of the quartiles. As is clear, the countries with
lower governance have much higher T&D losses.
Electricity theft and corruption appear to be closely
linked. It takes the connivance of power sector employees and even some politicians to perpetuate a system
whereby high T&D losses and theft continue to occur.

From a business perspective, electricity theft results in

economic losses to the utility. Some may argue that
large utilities providing essential services give poor
service, over-charge, make too much money anyway,
and, therefore, some theft will not break the company or
drastically affect its operations and protability. Others
looking at the same situation would argue that theft is a
crime and should not be allowed. An International
Utilities Revenue Protection Association has been
established to promote the detection and prevention of
power theftmainly for the nancial security of power
utility companies.
The consequences of theft in the worst case systems
are important to the viability of the services provided.
The combined losses (including non-payment of bills) in
some systems have severe impacts resulting in utilities
operating at a loss and must continually increase
electricity charges. Locked into a culture of inefciency
and corruption, the electricity utilities have difculty
delivering reliable service. Even in reasonably efcient
power systems, such as Malaysias Tenaga, power theft
accounts for losses of RM$500 million ($132 million)
annually (Malay Mail, 1999). For large systems a 1%
theft loss can be substantial. With sales of over $13
billion, 1% of theft for the Korea Electric Power
Corporation is over $130 million. Lovei and McKechnie
(2000) make a case that power theft impacts upon the
poor by perpetuating a system that benets the wealthy
and powerful. Power systems may also promote Grand
Theft by awarding lucrative contracts and monopolies
that lead the enrichment of favored individuals.
Indias power system is an illustration of a worst-case
situation. In constant turmoil, State Electricity Boards
(SEBs) have a high theft level and consumers do not pay
their bills. The SEBs seldom have prots and are heavily
subsidized for their losses (Smith, 1993). Only three
SEBs made a prot in 1996/97 and the combined
commercial losses were over 71 billion Rupees (about
$1.6 billion). The SEBs cannot pay their bills for power
purchased from the central government or IPPs nor for
plant equipment and the railways for coal delivery. The
whole system has been on the verge of nancial collapse

ARTICLE IN PRESS
T.B. Smith / Energy Policy 32 (2004) 20672076

for a decade. IPPs, especially foreign owned ones, are

reluctant to enter the power eld for fear that SEBs will
not be able to pay them for power supplied.

9. What can be done?

Electricity theft can never totally be eradicated in any
power system. In the very efcient systems of Japan,
Western Europe and North America effort has been
devoted to the technological and managerial methods
necessary to reduce theft to levels tolerable. Many of
these systems operate in a governance culture that
promotes organizational efciency and theft law enforcement. This does not mean that electricity consumers necessarily love their power company, but few
will try to steal electricity.
Power system strategies for dealing with theft vary.
Some organizations pay little attention to theft problems, perhaps hoping theft will disappear and not
become a public issue. Other power systems treat
electricity theft as highest priority.
The rst-step in electricity theft reduction is to
become knowledgeable about the theft problem. Few
detailed studies of power theft exist and the work of the
Prayas Energy Group (2002) in India provides many
insights.
Unless the nature and extent of power theft is known
in great detail, any attempts to deal effectively with the
problem are prone to fragmented and limited action that
have little over-all success. Therefore, power systems,
whether national or regional, should be encouraged to
initiate a detailed power theft analysis. The analysis
must go beyond conventional engineering and managerial frameworks and understand and explain why theft
occurs and what factors perpetuate theft. The information derived is essential to design an appropriate
strategy for dealing with theft.

10. Reducing power theft

Three methods for reducing power theft are identied
in this paper: technical/engineering, managerial, and
system change.
10.1. Technical/engineering methods
Electric power is not a new technology and innovations taking place enable very efcient systems to be
installed and maintained. Many power systems devote
inadequate resources and effort to transmission and
distribution systems and do not use the latest technologies. The investment necessary to reduce losses includes
upgrading power lines, transformers, information
technology monitoring systems, and installing and

2073

maintenance of modern metering systems that are at

the interface of the organization and the consumers of
the electricity.
Signicant technological advancement in metering
has occurred. Since much theft is from meter tampering,
it is important to replace old, easy to tamper-with
meters. New high-tech sealed meters that cannot be
altered in any way and can be read automatically are
costly, but can reduce theft when required of moderate
to heavy power users (see Arruda, 2000; Iyer, 2000;
Rajan, 1998). Szilvagyi (1999) makes a strong case that
the investment in high technology metering requires a
sound and complex infrastructure in place to make the
system work effectively.
10.2. Managerial methods
Electric power organizations are very large
entities that operate as bureaucracies even though
many are private sector organizations. Combining
strong technical improvements with an intelligent
and active anti-theft program may result signicant
improvements (see Ahmedabad Electricity Co. Ltd.,
2000).
Inspection and monitoring power users at regular
intervals is essential to reducing theft (Gower, 2000). In
Brazil, CEMIG had losses of $12 million. By spending
$2.1 million on tests and inspection, $6.2 million was
recovered (Arruda, 2000). The focus should be on areas
or facilities that have the greatest potential amount of
electricity theft in terms of electricity use. Studies have
shown that the wealthy steal power for residential use,
factories, and businesses (BRDC, 2000). More people
may be stealing power in urban slum areas, but the
amount of power is small by comparison. Yet inspection
often targets the poor of the community.
Singapores former Prime Minister Lee Kuan Yew
commented that corruption was a fact of life and in
Singapore it should not become a way of life. The
same comment could apply to electricity theft. Theft
may be prevalent in all power systems to varying degrees
as a fact of life. Clearly, some power systems appear
to be operating where electricity theft has become a
way of life.
Corruption is one of the most difcult problem areas
for electricity organizations because power theft occurs
with the connivance of employees of the power
organization. Increased investigation and surveillance
may provide opportunity for more corruption (Anuradha, 2000). Employees may even extort money from
electricity consumers not to disclose theft. It is
important to detect and prosecute corrupt power sector
employeesthis includes, if necessary, the ones at the
very top of the organization. Employees should be paid
adequately so that they will not have to resort to bribes
in order to support a family.

ARTICLE IN PRESS
2074

T.B. Smith / Energy Policy 32 (2004) 20672076

The organizational factor in the power industry is

important. Power utilities are very large, complex
organizations. By the number of employees it can be a
countrys largest organization. EGAT and the two
distribution agencies in Thailand have over 60,000
employees, Indonesias PLN has over 50,000. Tenaga
in Malaysia has 23,000 and WAPDA in Pakistan has
over 100,000. Nearly one million work in Indias state
electricity boards. Most of the tasks are routine and in
many organizations a bureaucratic culture is promoted
whether private or public enterprise.
Electricity utility employees must interface extensively
with the consumers of electricityin residences, factories and ofces. This allows street level decision
making to take place (Lipsky, 1980; Hudson, 1993).
Employees can exercise discretion by not reporting
infringements or may alter bills. Since the typical power
sector organization must operate at the consumer level,
employees are scattered throughout the far corners of
the country, making control and coordination from the
central ofce difcult. When the product delivered is a
scarce and essential commodity, as is electric power in
South Asian countries, employees can exercise considerable discretion. Routine allocation found in some power
systems becomes discretionary in others. For example,
who will get connected to power? When will the
connection be made? Where and when will power
blackouts take place? How much should the user pay
for power? These discretionary decisions can be for
sale by the employees. The organizations management
and employees thrive on power scarcity and there is little
incentive to increase supply or to operate a more
efcient or effective service.
The legal aspects of power theft have received
attention in some countries. Outdated laws treat theft
as a common crime. Several countries recently have
adopted laws governing power theft and treat it as a
special crime. The Andhra Pradesh amendments to the
Indian Electricity Act (1910) contains punishments from
6 months to 5 years imprisonment, nes of between 5000
to 50,000 Rupees, and depriving the thief of electric
power for up to 6 years. In Malaysia half-page ads
newspapers warn consumers of the illegality of power
theft with nes of up to RM 100,000 and imprisonment
of up to 5 years. The new laws make the punishment for
theft much easier to implement and the nes and
penalties imposed a deterrent to future theft.
The problem of arrears or non-payment is a difcult
one. Electricity is an essential commodity and a no pay,
no electricity policy may not be politically acceptable in
some countries. Disconnection also can be dangerous as
a World Bank (1999) study noted, In Albania,
consumers with guns y threatened to shoot the utility
ofcials who attempted to disconnect defaulting customers. The scope of this problem can be so serious that
the nancial viability of the organization is jeopardized.

Contracting the bill collection to a private agency may

promote some effectiveness in revenue collection. Providing alternative methods and places for bill payment
may also help. Some power systems have promoted prepaid cards as a method to ensure payment. However,
changing a culture of non-payment has no easy solutions
(Barnes, 2000; Lundin, 2001). In some cases those owing
the most money are government agencies, and collecting
can confront legal and political hurdles.
10.3. System change
In the systems where power theft is the highest,
electricity sector organizations are state owned and
managed enterprises. Some power sector state enterprises have operated with substantial efciency (in
Singapore, for instance), so one cannot argue a case
that the public sector is incapable of running services
effectively and efciently.
However, a case can be made that state owned and
operated enterprises are not managed as true businesses
and therefore do not try to optimize prots. The
organizations may be intertwined into the political and
bureaucratic structures and processes and there are few
incentives to reduce theft. In the Indian case, theft did
not slowly emerge, it has been around for many
decadesit is just that nothing was ever done about it.
Political leaders, power consumers and SEB managers
and employees have beneted from the system.
A world trend has been deregulation and the
transformation of public sector enterprises into the
private sector. In the past decade many power systems
were privatized and now operate as businesses with
shares traded on the stock exchanges (Bacon, 1999). The
total power sector is difcult to privatize into effective
private sector enterprises because transmission and
distribution are natural monopolies, and competition
is essential to spur businesses to be more efcient.
National and state level power systems have been
transformed in the past decade and the creation of an
independent regulatory commission for electricity has
been a common reform. The problem of how to deal
with technical and non-technical losses is a complex one
for the new commissions. The issues to grapple with
include setting levels of acceptable loss, whether
utilities should be allowed to pass on theft and other
inefciency costs to customers, and whether utilities
should be penalized if they do not achieve reductions in
T&D and theft.
The transformation of electric power systems into
more business-like enterprises means for many countries
the elimination of subsidies provided by the state that
kept electricity prices low for consumers. As prices in
poor countries rise to international levels, many consumers are trapped. Their own income is by local
standardsperhaps $2 to $5 per day, but their electricity

ARTICLE IN PRESS
T.B. Smith / Energy Policy 32 (2004) 20672076

charges are the same as for a customer in Los Angeles

who earns $80 per day. Under these conditions,
consumers may feel that there is no alternative but to
engage in electricity theft or not pay their bills.
Logic and theory suggests that private owned power
organizations will be more concerned with theft than
public sector organizations. Contrasting Malaysias
privatized system with Thailands public enterprise
system regarding electricity theft is interesting (Smith,
2003). Both systems have similar T&D losses of around
11%. In 1994 Malaysia divested Tenaga, the power
generation, transmission and distribution enterprise for
peninsular Malaysia. Government maintains majority
ownership, but its shares are traded on the Kuala
Lumpur stock exchange. Independent power producers
(IPPs) were permitted from the mid-1990s to produce
power and sell it to Tenaga for distribution. In the Thai
case, the EGAT is a public enterprise that generates and
transmits power to two large distribution public
enterprises, the Provincial Electricity Authority (PEA)
and the Metropolitan Electricity Authority (MEA).
Attempts to privatize Thai electricity have been discussed
for nearly 20 years, but the 32,000 member EGAT
employees union has vigorously opposed the change.
Electricity theft is not a big issue in Thailand because
EGAT, PEA and MEA appear to have no concerted
effort to deal with it. The enterprises make sufcient
prots to keep the government happy and to provide the
employees with free electricity as well as a substantial
end of year bonus in EGAT equal to about US$1000 per
employee. By contrast, in Malaysia, Tenaga has had to
display its operations and prots to the public and
shareholders. The recent economic crisis severely dented
Tenagas protability. Low prots affect the stock
market price of shares. Forced to run efciently, Tenaga
management turned, in a very serious way, to the
reduction of power theft that causes losses of M$500
million a year.
Caution needs to be exercised about promoting
privatization as a panacea for the ills of inefciency.
The Orissa (India) electricity sector was privatized in
1996 with the corporatization of the Orissa State
Electricity Board, the establishment of the Grid
Corporation of Orissa to manage T&D of electricity
and the Orissa Electricity Regulatory Commission to
regulate the system. The record shows uneven improvement (see Dixit et al., 1998). Power tariffs went up by
76%, T&D losses soared to 45%, and revenue collection
was only at 54% of those billed (Dhume, 1999).

2075

turmoil. Investment in improving the system and adding

additional capacity cannot be undertaken, loans and
payments cannot be met, and the consumer faces
increased electricity charges. Even in efcient systems,
theft losses can account for millions of dollars each year
in lost revenue.
Electricity theft in its various forms can be reduced
and kept in check only by the strong and assertive action
of power sector organizations. The strategy and the
action should be based upon a thorough understanding
of the specic nature of the theft problem. A strong case
can be made that each power system (including
consumers attitudes and behavior) has its own unique
qualities and only by knowing the system and the
problem can effective solutions be designed and
implemented. Since a high level of power theft is linked
with corruption, the analysis cannot be conned to
technical and managerial perspectives and needs to be
multi-disciplinary in approach.
Theft as an activity in some systems is closely
intertwined with governance and with the social,
economic and political environment. Corruption and
electricity theft thrives off each other. In an overall
culture of corruption as a way of life, electricity theft can
be reduced to moderate levels by technical/engineering
methods. But it is an uphill battle to reduce the
electricity theft rate drastically as long as extensive
corruption continues.
Reduction in power theft and keeping it within
reasonable bounds is more likely to be successful in
systems with a good governance culture. This is because
the theft reduction mechanisms nd a friendly environment for initiation and implementation. As part of
generating and sustaining good governance in communities, electric power systems have the opportunity to
take the lead in promoting sound corporate governance.
The technological innovations make this task easier
should the managerial skills and desire exist. Electric
power systems can be restructured to make power sector
organizations operate in competitive environments
where efciency and effectiveness in service delivery
are both virtues and necessities.
Acknowledgements
The author thank Dr. Ron Watts for assistance and
comments on this paper. This paper is an abridged and
revised version of a report, Electricity Theft: Understanding a Growing Problem prepared for InovaTech
Australasia, Ptv. Ltd.

11. Conclusions
The evidence points to the increasing levels of power
theft in many countries and the nancial losses for some
systems are so immense that the utility is in nancial

References
Ahmedabad Electricity Co. Ltd., 2000. The tiger has teeth: vigilance and
automation slash energy theft in India. Metering International 2.

ARTICLE IN PRESS
2076

T.B. Smith / Energy Policy 32 (2004) 20672076

Anuradha, G., 2000. KPTC vigilance ofcer suspended. The Asian

Age, 16 March.
Arruda, L.F., 2000. Accurate metering raises protsa holistic
revenue protection program. Metering International 2.
Bacon, R., 1999. A scorecard for energy reform in developing
countries. Public Policy for the Private Sector, April.
Barnes, J., 2000. Factors that Result in a Culture of Non-Payment.
Paper presented to the Fourth Annual South Africa Revenue
Protection Conference, Pietermaritzburg.
BRDC, 2000. Energy Loss Reduction Concepts. BRDC, Houston,
http://www.brdc.com/energyloss.htm
Campbell, K.J., 1999. Revenue protection: the good, the bad and the
ugly. Paper presented at the Third Annual South Africa Revenue
Protection Conference.
Culwell, J.J., 2001. Research study quanties energy theft losses.
Metering International Issue 1.
Dixit, S., Sant, D., Wagle, S., 1998. Regulation in the WB-Orissa
model: cure worse than disease. Economic and Political Weekly 33
(17), 112.
Dhume, S., 1999. Crossed wires. Far Eastern Economic Review 12 (8),
4041.
East European Energy Report, 1999. Elmu amnesty for power thieves
fails to net big sh. Financial Times, 27 January.
Flavin, C., Lenssen, N., 1994. Reshaping the electric power industry.
Energy Policy 22 (12), 10291044.
Gower, A.H., 2000. Non-technical losses: how much are you losing?
Paper presented at the Fourth Annual South Africa Revenue
Protection Association Conference.
Hudson, B., 1993. Michael Lipsky and street-level bureaucracy: a
neglected perspective. In: Hill, M. (Ed.), The Policy Process: A
Reader. London, Harvester-Wheatsheaf, pp. 386398.
Indian Express, 1998. Power theft accused absconding: one held, 9
September.
Iyer, A.G., 2000. Intelligent Metering Required, Indiainfoline, http://
www.indiainfoline.com.
Jakarta Post, 2000. Military Owes Rp. 23 billion to PLN, 21 March.
Kaufmann, D., Kraay, A., Zoido-Lobaton, P., 1999. Governance
Matters. Policy Research Working Paper 2196, Washington, DC,
World Bank Institute.
Kaufmann, D., Kraay, A., Zoido-Lobaton, P., 2003. Governance
Matters II: updated indicators for 2000/01, Washington, DC,
World Bank Institute Governance, Regulation and Finance
Division.
Lipsky, M., 1980. Street Level Bureaucracy: Dilemmas of the
Individual in Public Services. New York, Russell Sage Foundation.
Lovei, L., McKechnie, A., 2000. The costs of corruption for the
poorthe energy sector. Energy Services for the Worlds Poor
Washington, DC, World Bank (Chapter 8).

Lundin, B.L.V., 2001. The bottom line: getting customers to pay.

Energy Customer Management, 15 June.
Malay Mail, 1999. Tenaga out to short-circuit electricity thefts, 1
January.
Mkhwanazi, X.H., 1999. Electricity as a birthright and the problem of
non-payment. Paper presented at the Third Annual South Africa
Revenue Protection Conference.
News International, 2000. KESC MD links change in tariff to recovery
position, 19 July.
New Straits Times, 1999. Six suspected to be behind electricity thefts
arrested, 17 January.
National Economic Development Authority, 1998. Medium Term
Philippine Development Plan, 19982003 Manila, Government of
the Philippines.
Nesbit, B., 2000. Thieves lurkthe sizeable problem of stolen
electricity. Electrical World T&D, September/October.
Patterson, W., 2001. Transforming Electricity: the Coming
Generation of Change. Royal Institute of International Affairs,
London.
Priatna, D.S., 1999. Theft of electricity in Indonesia. Ph.D. Dissertation, University of Wisconsin, Madison.
Prayas Energy Group, 2002. HT Energy Audit: The Crucial Starting
Point for Curbing Energy Loss. Pune, Prayas Occasional Report 1/
2002.
Rajan, A.V., 1998. Moving beyond AMR. Metering International 2.
Rizvi, M., 2000. Pakistan gets switched on. VSO Orbit: Development
Magazine on Global Issues, Number 71.
Smith, T.B., 1993. Indias electricity crisis: why do the lights go out?
Asian Survey 33 (4), 376392.
Smith, T.B., 2003. Privatising electric power in Malaysia and Thailand:
politics and infrastructure development policy. Public Administration and Development 23 (2).
The Star, 1998. 59 own up to tampering with meters, 12 December.
Sullivan, K., 2002. Power pirates leave Mexico City in the dark.
Washington Post, 20 January.
Szilvagyi, M.J., 1999. AMRthe cure for energy theft? Metering
International 3.
Tacis, 1998. Improving Residential Electricity Services. Tacis Technical Dissemination Project (http://www.tacisinfo.ru).
Transparency International, 1999. Corruption in Public Sector
Departments. Dacca, TI Bangladesh Documents.
Tripathi, A., 2000. Lucknow steals power worth Rs 100 crore! The
Economic Times, 16 March.
World Bank, 1999. Non Payment in the Electricity Sector in Eastern
Europe and the Former Soviet Union, World Bank Technical
Paper No. 423, Washington, DC, World Bank.
World Bank, 2003. 2003 World Development Indicators, Washington,
DC, World Bank.